[24/95] ilog2: improve ilog2 for constant arguments

Message ID	20201216044337.Kjv9OiIPi%akpm@linux-foundation.org (mailing list archive)
State	New, archived
Headers	show Return-Path: <SRS0=WzRt=FU=kvack.org=owner-linux-mm@kernel.org> DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 92ED52313C Date: Tue, 15 Dec 2020 20:43:37 -0800 From: Andrew Morton <akpm@linux-foundation.org> To: akpm@linux-foundation.org, christophe.leroy@csgroup.eu, jakub@redhat.com, linux-mm@kvack.org, mm-commits@vger.kernel.org, peterz@infradead.org, rdunlap@infradead.org, torvalds@linux-foundation.org Subject: [patch 24/95] ilog2: improve ilog2 for constant arguments Message-ID: <20201216044337.Kjv9OiIPi%akpm@linux-foundation.org> In-Reply-To: <20201215204156.f05ec694b907845bcfab5c44@linux-foundation.org> User-Agent: s-nail v14.8.16 Sender: owner-linux-mm@kvack.org Precedence: bulk
Series	[01/95] mm: fix a race on nr_swap_pages \| expand [01/95] mm: fix a race on nr_swap_pages [02/95] mm/memory_hotplug: quieting offline operation [03/95] alpha: replace bogus in_interrupt() [04/95] procfs: delete duplicated words + other fixes [05/95] proc: provide details on indirect branch speculation [06/95] proc: fix lookup in /proc/net subdirectories after setns(2) [07/95] fs/proc: make pde_get() return nothing [08/95] asm-generic: force inlining of get_order() to work around gcc10 poor decision [09/95] kernel.h: split out mathematical helpers [10/95] kernel/acct.c: use #elif instead of #end and #elif [11/95] include/linux/bitmap.h: convert bitmap_empty() / bitmap_full() to return boolean [12/95] bitmap: remove unused function declaration [13/95] lib/test_free_pages.c: add basic progress indicators [14/95] lib/stackdepot.c: replace one-element array with flexible-array member [15/95] lib/stackdepot.c: use flex_array_size() helper in memcpy() [16/95] lib/stackdepot.c: use array_size() helper in jhash2() [17/95] lib/test_lockup.c: minimum fix to get it compiled on PREEMPT_RT [18/95] lib/list_kunit: follow new file name convention for KUnit tests [19/95] lib/linear_ranges_kunit: follow new file name convention for KUnit tests [20/95] lib/bits_kunit: follow new file name convention for KUnit tests [21/95] lib/cmdline: fix get_option() for strings starting with hyphen [22/95] lib/cmdline: allow NULL to be an output for get_option() [23/95] lib/cmdline_kunit: add a new test suite for cmdline API [24/95] ilog2: improve ilog2 for constant arguments [25/95] lib/string: remove unnecessary #undefs [26/95] lib: string.h: detect intra-object overflow in fortified string functions [27/95] lkdtm: tests for FORTIFY_SOURCE [28/95] string.h: add FORTIFY coverage for strscpy() [29/95] drivers/misc/lkdtm: add new file in LKDTM to test fortified strscpy [30/95] drivers/misc/lkdtm/lkdtm.h: correct wrong filenames in comment [31/95] lib: cleanup kstrto() usage [32/95] lib/lz4: explicitly support in-place decompression [33/95] bitops: introduce the for_each_set_clump macro [34/95] lib/test_bitmap.c: add for_each_set_clump test cases [35/95] gpio: thunderx: utilize for_each_set_clump macro [36/95] gpio: xilinx: utilize generic bitmap_get_value and _set_value [37/95] checkpatch: add new exception to repeated word check [38/95] checkpatch: fix false positives in REPEATED_WORD warning [39/95] checkpatch: ignore generated CamelCase defines and enum values [40/95] checkpatch: prefer static const declarations [41/95] checkpatch: allow --fix removal of unnecessary break statements [42/95] checkpatch: extend attributes check to handle more patterns [43/95] checkpatch: add a fixer for missing newline at eof [44/95] checkpatch: update __attribute__((section("name"))) quote removal [45/95] checkpatch: add fix option for GERRIT_CHANGE_ID [46/95] checkpatch: add __alias and __weak to suggested __attribute__ conversions [47/95] checkpatch: improve email parsing [48/95] checkpatch: fix spelling errors and remove repeated word [49/95] checkpatch: avoid COMMIT_LOG_LONG_LINE warning for signature tags [50/95] checkpatch: fix unescaped left brace [51/95] checkpatch: add fix option for ASSIGNMENT_CONTINUATIONS [52/95] checkpatch: add fix option for LOGICAL_CONTINUATIONS [53/95] checkpatch: add fix and improve warning msg for non-standard signature [54/95] checkpatch: add warning for unnecessary use of %h[xudi] and %hh[xudi] [55/95] checkpatch: add warning for lines starting with a '#' in commit log [56/95] checkpatch: fix TYPO_SPELLING check for words with apostrophe [57/95] checkpatch: add printk_once and printk_ratelimit to prefer pr_<level> warning [58/95] fs/nilfs2: remove some unused macros to tame gcc [59/95] kdump: append uts_namespace.name offset to VMCOREINFO [60/95] rapidio: remove unused rio_get_asm() and rio_get_device() [61/95] gcov: remove support for GCC < 4.9 [62/95] gcov: fix kernel-doc markup issue [63/95] bfs: don't use WARNING: string when it's just info. [64/95] relay: remove unused buf_mapped and buf_unmapped callbacks [65/95] relay: require non-NULL callbacks in relay_open() [66/95] relay: make create_buf_file and remove_buf_file callbacks mandatory [67/95] relay: allow the use of const callback structs [68/95] drm/i915: make relay callbacks const [69/95] ath10k: make relay callbacks const [70/95] ath11k: make relay callbacks const [71/95] ath9k: make relay callbacks const [72/95] blktrace: make relay callbacks const [73/95] kernel/resource.c: fix kernel-doc markups [74/95] ubsan: remove redundant -Wno-maybe-uninitialized [75/95] ubsan: move cc-option tests into Kconfig [76/95] ubsan: disable object-size sanitizer under GCC [77/95] ubsan: disable UBSAN_TRAP for allconfig [78/95] ubsan: enable for all*config builds [79/95] ubsan: remove UBSAN_MISC in favor of individual options [80/95] ubsan: expand tests and reporting [81/95] kcov: don't instrument with UBSAN [82/95] lib/ubsan.c: mark type_check_kinds with static keyword [83/95] reboot: refactor and comment the cpu selection code [84/95] reboot: allow to specify reboot mode via sysfs [85/95] reboot: remove cf9_safe from allowed types and rename cf9_force [86/95] reboot: allow to override reboot type if quirks are found [87/95] reboot: hide from sysfs not applicable settings [88/95] fault-injection: handle EI_ETYPE_TRUE [89/95] lib/lzo/lzo1x_compress.c: make lzogeneric1x_1_compress() static [90/95] apparmor: remove duplicate macro list_entry_is_head() [91/95] mm: unexport follow_pte_pmd [92/95] mm: simplify follow_pte{,pmd} [93/95] mm: fix some spelling mistakes in comments [94/95] mmap locking API: don't check locking if the mm isn't live yet [95/95] mm/gup: assert that the mmap lock is held in __get_user_pages()

Message ID

20201216044337.Kjv9OiIPi%akpm@linux-foundation.org (mailing list archive)

State

New, archived

Headers

DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 92ED52313C
Date: Tue, 15 Dec 2020 20:43:37 -0800
From: Andrew Morton <akpm@linux-foundation.org>
To: akpm@linux-foundation.org, christophe.leroy@csgroup.eu,
 jakub@redhat.com, linux-mm@kvack.org, mm-commits@vger.kernel.org,
 peterz@infradead.org, rdunlap@infradead.org,
 torvalds@linux-foundation.org
Subject: [patch 24/95] ilog2: improve ilog2 for constant arguments
Message-ID: <20201216044337.Kjv9OiIPi%akpm@linux-foundation.org>
In-Reply-To: <20201215204156.f05ec694b907845bcfab5c44@linux-foundation.org>
User-Agent: s-nail v14.8.16
Sender: owner-linux-mm@kvack.org
Precedence: bulk

Series

[01/95] mm: fix a race on nr_swap_pages | expand

Commit Message

Andrew Morton Dec. 16, 2020, 4:43 a.m. UTC

From: Jakub Jelinek <jakub@redhat.com>
Subject: ilog2: improve ilog2 for constant arguments

As discussed in https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97445 the
const_ilog2 macro generates a lot of code which interferes badly with GCC
inlining heuristics, until it can be proven that the ilog2 argument can or
can't be simplified into a constant.

It can be expressed using __builtin_clzll builtin which is supported by
GCC 3.4 and later and when used only in the __builtin_constant_p guarded
code it ought to always fold back to a constant.  Other compilers support
the same builtin for many years too.

Other option would be to change the const_ilog2 macro, though as the
description says it is meant to be used also in C constant expressions,
and while GCC will fold it to constant with constant argument even in
those, perhaps it is better to avoid using extensions in that case.

[akpm@linux-foundation.org: coding style fixes]
Link: https://lkml.kernel.org/r/20201120125154.GB3040@hirez.programming.kicks-ass.net
Link: https://lkml.kernel.org/r/20201021132718.GB2176@tucnak
Signed-off-by: Jakub Jelinek <jakub@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 include/linux/log2.h |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

--- a/include/linux/log2.h~ilog2-improve-ilog2-for-constant-arguments
+++ a/include/linux/log2.h
@@ -156,7 +156,8 @@  unsigned long __rounddown_pow_of_two(uns
 #define ilog2(n) \
 ( \
 	__builtin_constant_p(n) ?	\
-	const_ilog2(n) :		\
+	((n) < 2 ? 0 :			\
+	 63 - __builtin_clzll(n)) :	\
 	(sizeof(n) <= 4) ?		\
 	__ilog2_u32(n) :		\
 	__ilog2_u64(n)			\

[24/95] ilog2: improve ilog2 for constant arguments

Commit Message

Patch